Documentation Index
Fetch the complete documentation index at: https://docs.tokenlab.sh/llms.txt
Use this file to discover all available pages before exploring further.
TokenLab 实施速率限制以确保公平使用和平台稳定性。限制会因账户层级而异。
速率限制层级
| 层级 | 请求数/分钟 | 描述 |
|---|
| User | 1,000 | 所有账户的默认层级 |
| Partner | 3,000 | 适用于集成合作伙伴 |
| VIP | 10,000 | 适用于高吞吐量用户 |
速率限制响应
当你超出速率限制时，API 会返回 429 状态码，并附带一个 Retry-After header，用于指示在重试前需要等待多长时间。
超出速率限制
当你超出限制时，你将收到一个 429 响应：
{
"error": {
"message": "Rate limit exceeded. Please retry later.",
"type": "rate_limit_exceeded",
"code": "rate_limit_exceeded"
}
}
Retry-After header：
Retry-After: 60 # Seconds to wait before retrying
处理速率限制
指数退避
为自动重试实现指数退避：
import time
from openai import OpenAI, RateLimitError
client = OpenAI(
api_key="sk-your-api-key",
base_url="https://api.tokenlab.sh/v1"
)
def make_request_with_backoff(messages, max_retries=5):
for attempt in range(max_retries):
try:
return client.chat.completions.create(
model="gpt-4o",
messages=messages
)
except RateLimitError as e:
if attempt == max_retries - 1:
raise
wait_time = 2 ** attempt # 1, 2, 4, 8, 16 seconds
print(f"Rate limited. Waiting {wait_time}s...")
time.sleep(wait_time)
请求排队
对于高吞吐量应用，请实现请求队列：
import asyncio
from collections import deque
class RateLimitedClient:
def __init__(self, requests_per_minute=60):
self.rpm = requests_per_minute
self.interval = 60 / requests_per_minute
self.last_request = 0
async def request(self, messages):
# Wait if needed to respect rate limit
now = asyncio.get_event_loop().time()
wait_time = max(0, self.last_request + self.interval - now)
if wait_time > 0:
await asyncio.sleep(wait_time)
self.last_request = asyncio.get_event_loop().time()
return await self.client.chat.completions.create(
model="gpt-4o",
messages=messages
)
批处理
对于批量操作，请分批处理并添加延迟：
def process_batch(items, batch_size=50, delay=1):
results = []
for i in range(0, len(items), batch_size):
batch = items[i:i + batch_size]
for item in batch:
result = client.chat.completions.create(
model="gpt-4o",
messages=[{"role": "user", "content": item}]
)
results.append(result)
time.sleep(delay) # Pause between batches
return results
最佳实践
为相同请求缓存响应，以减少 API 调用。
更快的模型（如 gpt-5-mini）可支持更高的吞吐量。
升级你的层级
要申请升级层级：
- 登录你的 Dashboard
- 前往 Settings → Account
- 联系支持团队并说明你的使用场景
或者发送邮件至 support@tokenlab.sh，并提供：
